Question 1

What exactly are you licensing?

Accepted Answer

Operational business exhaust: email and chat threads, meeting transcripts, documents, tickets, CRM notes and project histories. It is used to train and evaluate enterprise AI systems. We are licensing scale, structure and realism — not any individual person's information.

Question 2

Why is our company's data valuable?

Accepted Answer

Leading AI models are short on real operational language — how actual businesses communicate, decide, sell, support customers and run projects. Public internet text doesn't capture that. Years of anonymized internal conversations and documents do, which is why model developers pay for it.

Question 3

Are you buying our data?

Accepted Answer

No. You license a copy for a defined purpose. You keep full ownership of your data and of your systems — nothing is purchased, transferred in ownership, or removed.

Question 4

What happens to our data after licensing?

Accepted Answer

It is used to train and evaluate leading AI models, and nothing else. Your corpus is not resold, syndicated or published.

Question 5

Is this legitimate? Who else does this?

Accepted Answer

Licensing proprietary data for AI training is now a standard practice among publishers, platforms and enterprises. Every engagement is governed by a written license agreement your counsel reviews before anything moves.

Question 6

Who qualifies?

Accepted Answer

US-based companies with more than 30 full-time team members that have been operating for more than 2 years. Longer histories and more systems generally mean a stronger valuation.

Question 7

What counts as a full-time desk employee?

Accepted Answer

Someone who works at a computer on operational business tasks — marketing, sales, finance, operations, support, engineering and similar. Field roles such as commercial truck drivers or warehouse workers don't count, and neither do contractors, freelancers, part-time or project-based workers.

Question 8

Will our headcount be verified?

Accepted Answer

Yes. After submission, payroll is verified in order to finalize your quote. Please report accurate numbers up front so your valuation doesn't change later.

Question 9

Do we need to include every system we use?

Accepted Answer

Yes. Companies must list all of their systems and cannot piecemeal the corpus. A complete, consistent picture of how the business operates is what makes the data valuable.

Question 10

Does it matter what language our data is in?

Accepted Answer

We ask what percentage of your data is in English. English-heavy corpora are most in demand today, but mixed-language data can still qualify.

Question 11

How much can we get paid?

Accepted Answer

Engagements typically range from $100,000 to $2 million, primarily driven by employee count.

Question 12

How is the price determined?

Accepted Answer

Price is driven by full-time desk headcount, the number of years covered, how many distinct systems are included, and total data volume. You get a written valuation range before you commit to anything.

Question 13

When do we get paid?

Accepted Answer

Payment is wired 30 to 60 days after delivery of the corpus.

Question 14

What does it cost us?

Accepted Answer

Nothing. There is no fee to submit, receive a valuation, or walk away. You are paid; you never pay us.

Question 15

Are we obligated to proceed after the valuation?

Accepted Answer

No. A valuation commits you to nothing. You only proceed once you've reviewed the number and the license agreement.

Question 16

Is any personally identifiable information included?

Accepted Answer

No. Every corpus goes through a multi-pass anonymization pipeline designed to remove names, emails, phone numbers, addresses, government IDs, payment details and credentials before delivery. You approve the redaction report before transfer is finalized.

Question 17

What about trade secrets or confidential business information?

Accepted Answer

You can flag categories, folders, channels or projects to exclude during scoping. Anything you mark as off-limits is excluded from the corpus.

Question 18

Who has access to the corpus after transfer?

Accepted Answer

Only research and evaluation teams, inside an isolated environment with audited access. Your corpus is never resold or passed to a third party.

Question 19

Do we lose access to our own data?

Accepted Answer

Never. You are licensing a copy. Your systems, retention policies and access controls are untouched.

Question 20

Can our security team review the process?

Accepted Answer

Yes — and we encourage it. Security and legal reviews are a normal part of every engagement. Reach us at info@datacorpus.ai.

Question 21

How long does the whole process take?

Accepted Answer

Most engagements close in 30 to 60 days: intake, valuation call, legal, anonymization review, transfer, then payment.

Question 22

What are the steps?

Accepted Answer

1) Submit a short intake with your company details and systems. 2) Get a valuation call and a written range. 3) Sign the license agreement. 4) Data is exported and anonymized; you approve the redaction report. 5) Transfer completes. 6) You're paid.

Question 23

How much work is this for our team?

Accepted Answer

Very little. Typically a technical lead grants read access to the relevant systems and joins a short kickoff. We handle export, processing and anonymization.

Question 24

How accurate do our data-size numbers need to be?

Accepted Answer

Accuracy is very important. Please look up the real figure in each system's admin console or usage report rather than estimating — it directly affects your quote.

Question 25

What systems can you take data from?

Accepted Answer

Google Workspace, Microsoft 365, Outlook, SharePoint, OneDrive, Slack, Teams, Zoom, Dropbox, Box, Notion, Confluence, Jira, HubSpot, Salesforce, Zendesk, GitHub, Snowflake, S3, Azure Blob and most other major business software.

Question 26

What if we use a system that isn't listed?

Accepted Answer

Add it at the bottom of the systems section in the intake form. We support most business software.

Question 27

What kinds of data are most valuable?

Accepted Answer

Long-running, high-volume, conversational data — email, chat, meeting transcripts, tickets and CRM notes — especially across many years and multiple teams.

Question 28

Does old or archived data count?

Accepted Answer

Yes. Historical data is often the most valuable part of a corpus. Archives sitting in cold storage are frequently worth more than recent data.

Straight answers.

The basics

Eligibility

Pricing & payment

Privacy & security

Process & timeline

Systems & data

Still deciding? Get the number first.

Have a specific question?